We seek renewal of our NLM Training Program in Biomedical Informatics (NLMTP), which the NLM has supported for twenty years. The program, evolving as biomedical informatics (BMI) itself has evolved, has consistently produced outstanding trainees. In this renewal, we address new challenges to BMI arising from digital information streaming from innumerable sensors, instruments and simulations, which are outrunning our capacity to accumulate, organize and analyze it. Advances in BMI are needed: 1) to make sense of datasets that may be massive, heterogeneous or even contain errors, 2) to cull from them important insights into fundamental problems of biomedicine and 3) to convey information in ways readily understood by researchers and clinicians. These are fundamental BMI problems, which we will address in the NLMTP renewal. We will emphasize health care/clinical informatics (HC), translational bioinformatics (TB) and clinical research informatics (CR), areas in which our trainees will contribute to progress needed to gain the full wealth of knowledge embedded in genomic, proteomic, genetic, epidemiological, and clinical data and a full return on our substantial investments in health information technology. Our trainees will bring to each of these BMI domains what we call data-driven inference and decision making (DIDM) by which we mean the use computer programs to seek associations in databases whose complexity hides such relations from even expert humans and cognitive informatics and decision making to make such discovered patterns intelligible to humans. Our pre-doctoral trainees will be selected from graduate students who have completed one year of study at one of six participating institutions. NMLTP training will combine DIDM and BMI specific courses and research to engage them with the range of opportunities in HC, TB and CR suggested above. We also will involve pre- and postdoctoral trainees in a set of courses and activities to facilitate their integration into the BMI community. Over many years, our NLMTP has brought computation, mathematics, statistics, simulation and advanced imaging to bear on biomedical problems that were before deemed intractable. Now our researchers are deploying powerful new computational methodologies for data exploration on an unprecedented scale that offer exciting opportunities for the training to prepare a new generation of BMI professionals. We are requesting 9 pre-doctoral and 6 postdoctoral positions for this renewal.